dr. richard moore - illumina · dr. richard moore sequencing group leader ... assemble viral reads...

© BC Cancer Agency Branch

Sequencing Cancer Genomesand Transcriptomes

Dr. Richard Moore

Sequencing Group Leader

Genome Sciences Centre, BC Cancer Agency


Capacity at the GSC

• 11 Illumina GA II• 2 AB SOLiD• 8 3730 xls

• Paired and Single end reads• Production 51 bases, Tech D upto 101

• Library TypesWhole Genome Shotgun MREmiRNA Small RNARNA seq / WTSS MeDIPChIP SAGEBisulfite


GSC Illumina Yield by Quarter

0

100,000,000,000

200,000,000,000

300,000,000,000

400,000,000,000

500,000,000,000

600,000,000,000

2007 2007 2007 2007 2008 2008 2008 2008 2009

Raw

bases

Total 1.6 Trillion bases

Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1


GSC Utilisation of UHTS technology

• 26 UHTS publications– Including Genome research and Nature Methods

• 1.6 trillion bases produced• >1200 libraries


Examples of Current Projects•Lobular Breast Cancer•Adenocarcinoma - personal genomics•Follicular Lymphoma

•Diffuse Large Bcell Lymphoma•Ovarian Cancer•Lung Cancer•Acute Lymophoblastic Leukaemia•Neuroblastoma•Prostate Cancer•Oligodendrioma•Colo-rectal Carcinoma


Cancer prevalenceCancer prevalence

Females 39.3 2.5 24.1 4.2

Males 44.5 2.2 28.5 3.5

Life time probability of:

developing dying % one in: % one in:

Canadian Cancer Society Statistics, 2008


Why sequence cancer genomesWhy sequence cancer genomesand and transcriptomestranscriptomes??

• To identify cancer driver mutations and pathways.– Challenges: passengers vs. drivers, undiscovered SNPs, changes selected for in

culture, genetic heterogeneity of cell populations…

• To identify targets for development of new therapies.

• To improve diagnostic precision and prognostic accuracy.– E.g. “breast cancer” describes several diseases.

• To match patients to treatments.– Optimize treatment modalities based on individual genes and genomes

• To understand differences in treatment response.– Outright failure

– Remission / relapse (treatment resistance).

• To manage treatment failure.– Alternative existing therapies?

7


Next-generation sequencing ofNext-generation sequencing ofcancer patient samplescancer patient samples

• Depth-dependent sensitivity to rare events; e.g.genomic changes in cellular sub-populations (tumorinitiating cells?).– Enabled by massively redundant “single molecule”

sequencing, in contrast to technologies thatintegrate signal from many molecules.

8

+/+

+/+ +/+

+/+

+/+

+/+

+/++/+ +/+

+/-

G G G

01 - 8


Lobular Breast Cancer


44.1638.9438.7Mean read length (bp)

306,626,95021,613,16698,727,198Unaligned reads

88.3267.7991.31Percent reads aligned canonically

439,409,649109,093,616721,929,588Canonically aligned reads

6.776N/A9.809Estimated depth(non-gap regions)

0.0190.0130.015Estimated error rate

21.9716.26630.599Aligned nucleotides (Gb)

497,521,910160,919,484790,665,100Number of aligned reads

35.5117.10834.419Total nucleotides (Gb)

804,148,860182,532,650889,392,298Total number of reads

WGSS-PRIWTSS-PEWGSS-PE

Samples and librariesSamples and libraries


Summary

•Alternate spicing shown in the estrogen signalling pathway

•25 NS somatic mutations

•16 only in the metastaic tumour

•2 RNA editing somatic NS mutations

•Novel amplicon region in the Insulin receptor


Follicular Lymphoma

BAC re-sequencing of large scalerearrangements


Summary of validated large-scale somatic events

We have identified and validated 52 somatic and 66 germline breakpoints, corresponding to 38 and 41 distinct events. Examples of somatic mutations include:

KITSequence additions201374q12RNG

AP3M1 and BTAF1Sequence additions1918,12610q22.2-10q23.32DUP

OPRM1Microhomology21806q25.2INV

RFTN1Blunt21NAt(1;3)(q23.3;p24.3)TRX

CDK5Microhomology20NAt(6;7)(q15;q36.1)TRX

AbParts, IGHSequence additions1939514q32.33DEL

AbParts, IGLSequence additions6, 17112-86122q11.22DEL

AbParts, IGKAll features8, 13, 16, 17193-4122p11.2DEL

LPAR3, MCOLN2, MCOLN3, WDR63,SYDE2, C1orf52, BAG and last exon ofBCL10Microhomology85581p22.3DEL

PTPRDBlunt102519p24.1DEL

CDKN2AMicrohomology161249p21.3DEL

IGH and BCL2Sequence additions6, 8, 10, 13, 16NAt(14;18)(q32;q21)TRX

SORCS1Microhomology10NAt(10;12)(q25.1;q23.1)TRX

BCL6 and ST6GAL1All features14, 21718/7223q27.3INV

Affected gene(s)Junction featurePatient IDEvent

size (kb)Chromosome band(s)Eventtype

DEL, Deletion; TRX, Translocation; INV, Inversion; DUP, Duplication; RNG, Complex rearrangement; NA, Not applicable


Fork Stalling and Template Switching – a modelfor nucleotide additions at breakpoint junctions

Cell. 2007 Dec 28;131(7):1235-47


Concluding remarks

• To date we have identified 38 distinct somatic events from 20

primary follicular lymphoma patient tumour samples.

• 40% of the rearrangement events assayed are acquired in the

tumour samples.

• Having sequence level resolution of breakpoints is informing on

mechanistic insights.

• For the first time in cancer, to our knowledge, we have

demonstrated that sequence additions at rearrangement

breakpoints are consistent with the FoSTeS model of DNA

replication.


Personal cancer genomics:Personal cancer genomics:

Adenocarcinoma, pulmonarymetastases -

non-responsive to treatment (erlotnib)


QuestionsQuestions

• In this case (rare tumor, no standard chemotherapy

options), can next generation genome analysis provide

clues as to the expression and mutational profiles of

known drug targets?

• Can next generation genome analysis provide insight

into the apparent resistance of the tumor to erlotinib,

despite apparent amplification of EGFR?

19


Samples and librariesSamples and libraries

20

Type Source Amount Total reads% alignedto genome

Sequence yieldtotal

(base-pairs)

Sequence yieldaligned (base-

pairs)

RNA-seq Normal: Blood 10ug Total RNA 85,795,100 72.9 3,603,394,200 2,625,754,824GenomeShotgun Normal: Blood 10ug gDNA 40,793,289 83.7 1,713,318,138 1,434,873,300

RNA-seq lite(cDNA amp)

Cancer: LungBiopsy 10ng Total RNA 27,897,5099 69.2 11,716,954,158 8,104,150,782

RNA-seq lite(cDNA amp)

Cancer: LungBiopsy 10ng Total RNA 431,797,605 69.7 18,135,499,410 12,645,184,944

RNA-seq lite(RNA amp)

Cancer: LungBiopsy 2ng Total RNA 6,328,326 66.3 265,789,692 176,282,652

GenomeShotgun

Cancer: LymphNode FFPE 10ug gDNA 2,942,711,791 87.8 123,593,895,222 108,551,254,728

Totals: 3,786,401,210 159,028,850,820133,537,501,230

~84%


Data analysisData analysis

• Transcriptome libraries:– Digital gene expression profiling, comparing tumor read counts

across ~23 in-house WTSS libraries.– Mutation detection.

• Genome libraries:– Digital karyotyping (copy number analysis), comparing tumor

genome sequence to PB (“normal”) genome sequence and theYoruban sequence.

– Mutation detection.

• Integration with drug bank:– Relate mutations, copy number alterations, and gene

expression data to known drugs.

21


Drug BankDrug Bank

• www.drugbank.ca

• 4,408 drugs in drugbank for 4535 targets.

• 1,359 FDA-approved drugs for 1613 targets

• 189 "cancer" drugs for 291 targets.

22


Intersection of drug bank withIntersection of drug bank withgenomic features and cancergenomic features and cancer

pathwayspathways• 30 amplified genes in cancer pathways.• 76 deleted genes in cancer pathways.• ~400 up- and ~400 downregulated genes.

– 19 of these in known cancer pathways• Single base changes (transcriptome):

– 303 non-synonymous point mutations in 177 genes– 4 genes in cancer pathways

• Other:– 233 novel coding SNPs,– 25 genes affected by indels– 126 genomic deletions spanning exons for 83 genes– No intersections with known cancer pathways

23


Sunitinib was chosen because it targets the most number of expressed genes in the tumour. It inhibitsRET, PDGFRA, CSF1R.

Sorafenib is also able to inhibit a mutant RET and B-RAF, and was thus selected as secondary treatmentin case the patient did not respond or acquired resistance to sunitinib.

Proposed therapeutic interventions


EGFR RET TNFR

PI3K

AKT

mTOR

p70S6K

SHCGRB SOS

proliferation

BAD

apoptosis

PTENPIP3

PIP2Ras

Raf

MEK

ERK

PDGFR

eIf4E

Rb1

TRAF2

JNK

AP-1

amplified

LOH deleted

mutated

upregulated

downregulated

eIf4EBP

growth

TRAF1

Integrating pathways, mutationsand gene expression


Response to Sunitinib

After 29days on Eroltinib tumours increasedin size by >20%

After 28 days on Sunitinib tumour reducedin size by >20%


Infectious Agents and Cancer


Total RNA extraction

cDNA fragment library construction 150-300bp

Sequencing on IlluminaGA

Ribominus depletionPolyA selectionTumour Biopsy Or

Genomic Subtraction ForInfectious Agent Detection


10million to 100millionreads

Map unmapped readsto human genome

(hg18)

Map unmapped readsto infectious agent

database

Pool hits to sameinfectious agent andrank candidates by

number of hits and type

Assemble viral reads

All Reads

Transcriptome Subtracted reads

Genome subtracted

reads

Map to humantranscriptome

(including splicejunctions)

Infectious Agentspecific reads


De-novo assembly of unmapped readsreveals viral RNA and integration site

• HeLa is derived from acervical cancer positive forHPV infection

• De-novo assembly ofunmapped reads reveals HPV-derived sequences

• A few HPV contigs containhuman sequence (revealintegration site of virus)


AcknowledgementsAcknowledgements

31

Bilenky, MishaBirol, InancChuah, EricCorbett, RichardGriffith, ObiHirst, Martin & teamHolt, RobJones, SteveLi, YvonneMarra, MarcoMoore, Richard & teamMorin, RyanMorozova, OlenaPugh, TrevorRoscoe, Robyn & teamShah, SohrabThiessen, NinaVarhol, Richard & teamZhao, Yongjun &team

Cancer patients and research advocates

Aparicio, SamHuntsman, DavidJones, StevenLaskin, Janessa

Funding:Genome CanadaGenome British ColumbiaBC Cancer foundationCIHRNCI (USA)NCI (Canada)Western Economic Diversification

dr. richard moore - illumina · dr. richard moore sequencing group leader ... assemble viral reads...

Documents